Chile Earthquake

Update 9 | Summary
Posting Date: March 26, 2010, 11:45:00 AM

On Saturday, February 27, a magnitude 8.8 earthquake—the fifth largest earthquake since instrumental record keeping began— struck off the central coast of Chile. Buildings and infrastructure from Valparaíso to the region of Araucanía were impacted. The area is home to about 80% of the country’s population. A tsunami followed the M 8.8 event, severely affecting several stretches of the coastline. In little more than 24 hours, on February 28, AIR issued an industry insured loss estimate of between 2 billion USD (CLP 1,050 billion) and 8 billion USD (CLP 4,200 billion), a range that reflected the uncertainty in the earthquake’s event parameters. Since then, AIR scientists and engineers have analyzed updated event parameters and findings from AIR’s post-disaster survey of the region. In light of their findings, AIR has refined its industry insured loss estimates for property lines excluding infrastructure to between 5.5 billion USD and 8.5 billion USD. Revised Source Parameters The source parameters used in the initial posting were estimated based on subduction zone geometry, early faulting mechanism data from the US Geological Survey (USGS) and other agencies, and aftershock distributions from the first 12 hours following the earthquake. Over the next several days, various agencies updated (more than once) their estimates of co-seismic slip distributions and the radiation strength and pattern of displacement across the fault. The new information that has been used to update AIR’s loss estimates includes a month of aftershock data and co-seismic slip results from three research institutions: the USGS, the University of California at Santa Barbra, and the Earthquake Information Center at Tokyo University, in Japan. The co-seismic slip results were particularly critical, as they provided improved constraints on the location and geometry of the rupture plane, addressing length, width, dip angle, and depth. The event parameters selected for these new scenarios reflect the reduced uncertainty as new data has been gathered and analyzed. Field Survey Findings AIR’s updated loss estimates also benefit from post-disaster field survey observations. AIR engineers were on site in various regions of central and southern Chile beginning Friday, March 12, until Sunday, March 21. The key goal of the survey was to obtain a detailed assessment of damage severity and distribution. The team was made up of AIR Principal Engineers Dr. Guillermo Franco and Dr. Tao Lai, and AIR Senior Engineer, Guillermo Leiva. Dr. Franco also served as member of the Earthquake Engineering Research Institute (EERI) team. Due to this event’s extreme magnitude and relatively shallow depth, the potential for significant damage was high. However, modern buildings in Chile performed extremely well, in general, sustaining minimal structural damage in major cities including Santiago, Concepción and Constitución. Minor damage—the kind not visible except on close general inspection—was fairly common. This building performance clearly demonstrates the efficacy of Chile’s strict building code, which has made it an attractive market for the global insurance and reinsurance industry. The localized full or partial failures observed in modern concrete structures will provide the engineering community with new information that will inform code revisions in due course. As expected, most of the structural collapses were restricted to Chile’s historic adobe building stock. Modeled and Non-modeled Losses AIR's insured loss estimates reflect: • Insured physical damage to property (residential, commercial/ industrial, auto), both structures and their contents • Direct business interruption losses • Demand surge—the increase in costs of materials, services, and labor due to increased demand following a catastrophic event The loss estimates do not reflect: • Losses to uninsured properties • Losses to infrastructure • Indirect business interruption losses • Loss adjustment expenses • Losses from non-modeled perils, including tsunami and fire-following Structures on the coastline were vulnerable to the subsequent tsunami. AIR’s survey team observed significant devastation in coastal communities throughout the Bío-Bío region, where the tsunami arrived with varying depths, reaching more than 6 meters in some locations. However, property losses generated by the tsunami were largely restricted to uninsured structures, such as wood frame dwellings in flood plains within 200 meters of the shore. There were damages to the container harbor of Talcahuano and to industrial facilities located in coastal areas between Concepción and Constitución. At some of these industrial facilities, the tsunami damage will likely result in business interruption losses. However, it is expected that these losses will account for a relatively small fraction of the overall destruction from the earthquake. Infrastructure suffered considerable damage during the earthquake. Currently, the Chilean government estimates that losses to infrastructure could reach 1.5 billion USD. The Chilean government requires that infrastructure operating under a concession is covered by earthquake insurance. Clients should note that the AIR modeled loss estimates do not include insured losses from damage to infrastructure. A portion of the losses from the M 8.8 earthquake in Chile will be attributed to business interruption (BI). Direct BI is included in the AIR loss estimates. However, indirect BI losses will also result as a consequence of utility service interruption, damage to the transportation infrastructure and damage to dependent buildings and suppliers. Indirect BI losses are not included in AIR’s loss estimates for Chile. The new range of modeled loss estimates is posted on the Scenarios page of the ALERT website along with ground motion and loss maps. Please note that the assumed exchange rate is 1 USD = 525 CLP. The industry insured loss estimates incorporate adjustments to AIR’s Industry Exposure Database and AIR’s estimated insurance take-up rates for Chile. Given the significant economic growth in the country during the past few years, AIR believes industry insured exposures are now approximately 30% higher than those represented in the 2005 Industry Exposure Database accompanying the most recent release of the AIR Earthquake Model for Chile in 2006. CATRADER® users who would like to replicate this industry insured loss range should apply a Loss Modification Factor of 1.30 in the Program Options tab prior to running this analysis. AIR has also reviewed its estimated insurance take-up rates for Chile given that nearly all residential properties with a mortgage now also have insurance on the building, with most of these policies in or around the Santiago metropolitan area. AIR estimates that for CRESTA Zone 3, which includes Santiago, the take-up rate for residential buildings should be 30% versus 10%. While AIR is not making adjustments to take-up rates for any other CRESTA zones or lines of business, we caution that take-up rate estimates are a source of uncertainty and clients may wish to make adjustments based on their own knowledge. To reproduce these industry loss estimates in CATRADER, clients should adjust the Zone 3 take-up rate for residential buildings as noted above in addition to the loss adjustment factor for the increase in industry exposures. IMPORTANT: Clients running exposures for individual companies in CATRADER or CLASIC/2™ should NOT make either of the adjustments noted above. The ALERT scenarios are intended to provide AIR’s best estimate of the losses covered by the model without the need to make these adjustments. The adjustments are only needed for CATRADER users wishing to generate industry loss estimates. As clients review their estimated losses from this event, AIR recommends the following best practices to enhance the interpretation and communication of results: • Due to unique nature of this event clients should use the ALERT event set to estimate company losses rather than relying on similar events from the stochastic event catalogue • Always communicate the range of loss estimates produced by the event sets provided on the ALERT site, not just any single estimate or just the maximum; • Always make sure the loss results include demand surge; • Benchmark estimated losses against losses based on market share information and industry loss estimates when appropriate; • Present any known issues with exposure data quality that might affect the loss results; • Disclose any adjustments made to reflect non-modeled costs such as loss adjustment expenses, fire following, or tsunamis. We encourage clients to review the AIR Earthquake Model for Chile technical document on the Client Access section of the AIR website. Sections 7 and 8 contain relevant information about the AIR Industry Exposure Database for Chile and the impact of different construction classes and age bands on loss results. This information can help in understanding how individual company loss estimates compare to industry loss estimates.

Update 9 | Downloads
Posting Date: March 26, 2010, 11:45:00 AM